Imagen de Portada

The Shape of Overthinking: Backtracking Bursts in Long Reasoning Traces

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Rezazadeh, Navid, Davoodi, Arash Gholami
Formato:	Preprint
Publicado:	2026
Materias:	Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Vertex-Softmax: Tight Transformer Verification via Exact Softmax Optimization
por: Rezazadeh, Navid, et al.
Publicado: (2026)

Mitigating Overthinking through Reasoning Shaping
por: Song, Feifan, et al.
Publicado: (2025)

LLMs Are Not Intelligent Thinkers: Introducing Mathematical Topic Tree Benchmark for Comprehensive Evaluation of LLMs
por: Davoodi, Arash Gholami, et al.
Publicado: (2024)

Geometry-Aware Decoding with Wasserstein-Regularized Truncation and Mass Penalties for Large Language Models
por: Davoodi, Arash Gholami, et al.
Publicado: (2026)

The Danger of Overthinking: Examining the Reasoning-Action Dilemma in Agentic Tasks
por: Cuadron, Alejandro, et al.
Publicado: (2025)

Reconsidering Overthinking: Penalizing Internal and External Redundancy in CoT Reasoning
por: Hong, Jialiang, et al.
Publicado: (2025)

Emergent Search and Backtracking in Latent Reasoning Models
por: Cui, Jasmine, et al.
Publicado: (2026)

The First Impression Problem: Internal Bias Triggers Overthinking in Reasoning Models
por: Dang, Renfei, et al.
Publicado: (2025)

Innate Reasoning is Not Enough: In-Context Learning Enhances Reasoning Large Language Models with Less Overthinking
por: Ge, Yuyao, et al.
Publicado: (2025)

Backtracking for Safety
por: Sel, Bilgehan, et al.
Publicado: (2025)

ASTRO: Teaching Language Models to Reason by Reflecting and Backtracking In-Context
por: Kim, Joongwon, et al.
Publicado: (2025)

Mitigating Overthinking in Large Reasoning Language Models via Reasoning Path Deviation Monitoring
por: Guan, Weixin, et al.
Publicado: (2026)

The Evolution of Thought: Tracking LLM Overthinking via Reasoning Dynamics Analysis
por: Wei, Zihao, et al.
Publicado: (2025)

Don't Overthink it. Preferring Shorter Thinking Chains for Improved LLM Reasoning
por: Hassid, Michael, et al.
Publicado: (2025)

Reasoning or Overthinking: Evaluating Large Language Models on Financial Sentiment Analysis
por: Vamvourellis, Dimitris, et al.
Publicado: (2025)

Between Underthinking and Overthinking: An Empirical Study of Reasoning Length and correctness in LLMs
por: Su, Jinyan, et al.
Publicado: (2025)

Don't Overthink It: A Survey of Efficient R1-style Large Reasoning Models
por: Yue, Linan, et al.
Publicado: (2025)

Think How to Think: Mitigating Overthinking with Autonomous Difficulty Cognition in Large Reasoning Models
por: Liu, Yongjiang, et al.
Publicado: (2025)

BacktrackAgent: Enhancing GUI Agent with Error Detection and Backtracking Mechanism
por: Wu, Qinzhuo, et al.
Publicado: (2025)

The Shape of Reasoning: Topological Analysis of Reasoning Traces in Large Language Models
por: Tan, Xue Wen, et al.
Publicado: (2025)

Mitigating Overthinking in Large Reasoning Models via Manifold Steering
por: Huang, Yao, et al.
Publicado: (2025)

How Much Backtracking is Enough? Exploring the Interplay of SFT and RL in Enhancing LLM Reasoning
por: Cai, Hongyi James, et al.
Publicado: (2025)

Don't "Overthink" Passage Reranking: Is Reasoning Truly Necessary?
por: Jedidi, Nour, et al.
Publicado: (2025)

Mitigating Overthinking in Large Reasoning Models via Difficulty-aware Reinforcement Learning
por: Wan, Qian, et al.
Publicado: (2026)

From Token to Action: State Machine Reasoning to Mitigate Overthinking in Information Retrieval
por: Lee, Dohyeon, et al.
Publicado: (2025)

Reasoning Traces Shape Outputs but Models Won't Say So
por: Hao, Yijie, et al.
Publicado: (2026)

Precedent-Informed Reasoning: Mitigating Overthinking in Large Reasoning Models via Test-Time Precedent Learning
por: Wang, Qianyue, et al.
Publicado: (2026)

Backtracking When It Strays: Mitigating Dual Exposure Biases in LLM Reasoning Distillation
por: Wang, Bing, et al.
Publicado: (2026)

Step Back to Leap Forward: Self-Backtracking for Boosting Reasoning of Language Models
por: Yang, Xiao-Wen, et al.
Publicado: (2025)

Your Models Have Thought Enough: Training Large Reasoning Models to Stop Overthinking
por: Han, Jinyi, et al.
Publicado: (2025)

BadThink: Triggered Overthinking Attacks on Chain-of-Thought Reasoning in Large Language Models
por: Liu, Shuaitong, et al.
Publicado: (2025)

Missing Premise exacerbates Overthinking: Are Reasoning Models losing Critical Thinking Skill?
por: Fan, Chenrui, et al.
Publicado: (2025)

Hop, Skip, and Overthink: Diagnosing Why Reasoning Models Fumble during Multi-Hop Analysis
por: Yadav, Anushka, et al.
Publicado: (2025)

Correction with Backtracking Reduces Hallucination in Summarization
por: Liu, Zhenzhen, et al.
Publicado: (2023)

Concise Reasoning, Big Gains: Pruning Long Reasoning Trace with Difficulty-Aware Prompting
por: Wu, Yifan, et al.
Publicado: (2025)

Reinforcement Learning with Backtracking Feedback
por: Sel, Bilgehan, et al.
Publicado: (2026)

Backtracking Improves Generation Safety
por: Zhang, Yiming, et al.
Publicado: (2024)

TACT: Mitigating Overthinking and Overacting in Coding Agents via Activation Steering
por: Sui, Yuan, et al.
Publicado: (2026)

BEAP-Agent: Backtrackable Execution and Adaptive Planning for GUI Agents
por: Lu, Ziyu, et al.
Publicado: (2026)

Deep Backtracking Counterfactuals for Causally Compliant Explanations
por: Kladny, Klaus-Rudolf, et al.
Publicado: (2023)